ON SEISMOLOGICAL INVESTIGATION. 183 
Assuming that on a diagram we can measure within ‘25 mm., be- 
5 fe) ’ 
cause 
V runs at 5 mm. per minute, therefore we can read to within 3 seconds. 
M ” ‘ ”) ” 7 +B) 15 ” 
(= 
R ” Ry 5B) ” ” ” 45 ” 
; L 
D ” 6 ” ” ” ” 30 » 
By shaking M at known times, and comparing these times with times 
determined from the developed film, the difference between these is about 
half the expected error. Because this is probably true for the other 
instruments, the errors in 100 seconds may be, 
V_ 1:5 seconds or per cent. 
(a) 
R 22'5 
D 45 ” ” 
2 
Because time-intervals in V and M do not depend upon the elock 
driving the record-receiving surface, but are marked by an independent 
time-keeper, these errors should not exceed a small fraction of a second 
per hour. R and D do not share this advantage, the time being 
dependent upon a clock driving a drum or a broad film of bromide paper. 
As a time recorder D is like R. Should the rate be made greater, 
it might involve an increase in light-source. The time intervals might 
also be marked by an independent clock. 
V and M also present the advantage of yielding a diagram, the 
definition of which is much sharper than Rand D.  V is slightly better 
than M. 
The M clock, which, however, only drives a film 2 inches wide, is so 
arranged that it can be instantly altered to drive the paper at a rate of 
about 6 or 10 in. per day, which, when recording diurnal waves, is 
sufficiently quick. 
Equality in Adjustment (important).—If two or more similar instru- 
ments are not adjusted to have equal sensibility, they may commence to 
indicate with different phases of motion, and much of what is gained in 
the accuracy of the time scale is lost. 
M and equivalent of R can be adjusted to have a close similarity in 
sensibility, and this is probably true of D. 
With V, which writes by the friction of a pointer on a smoked surface, 
we have no experience, but from experience with its equivalent and a 
large experience with ordinary seismographs writing upon smoked surfaces, 
it seems likely that there would be great difficulty in obtaining equal 
sensibility, especially with instruments which were not side by side. 
Even if absolute equality is attainable with a group of instruments, it 
should be remembered that instruments further from the epicentre will 
necessarily indicate a later phase of the movement than those close to it. 
Sensibility.— All types record long period wave motion. 
V gives an open diagram for movements, the period of which is not 
less than five seconds—that is of preliminary tremors at a distance from 
their origin. R and M show the presence of these, but of D we have no 
experience. 
R and D give diagrams of large amplitude, but V has the best 
definition. 
